T*: Re-thinking Temporal Search for Long-Form Video Understanding
Comments Accepted by CVPR 2025; A real-world long video needle-in-haystack benchmark; long-video QA with human ref frames
作者
Computer Vision
Comments Accepted by CVPR 2025; A real-world long video needle-in-haystack benchmark; long-video QA with human ref frames
机构 * Stanford University(斯坦福大学)
Comments 9th Conference on Robot Learning (CoRL 2025), Seoul, Korea. Project website: https://behavior-robot-suite.github.io/